1. Home
2. Reference documentation
3. Fastly API
4. Metrics and stats
5. Domain metrics

Domain Inspector provides a JSON-formatted historical API for visibility into responses delivered from your domains by Fastly. The API offers a standardized set of data to report in minutely, hourly, or daily granularity on every domain response, byte, and status code received by a specified service.

Query options

When the start or and parameters specify a date without a time, the time component will be set to 00:00. To specify a different time, use a Unix timestamp or an expression such as start=2021-01-01%2008:00&end=2021-01-02%2008:00

If start is omitted, the time range will start at a point determined by the value of downsample:

• ?downsample=day: one month ago
• ?downsample=hour: one day ago
• ?downsample=minute: 30 minutes ago

If you define values for all of the start, end, and downsamples parameters you can accidentally ask for a very large data set.

Examples

Returns domain response counts for service abc123, summed by day, for the first week of September 2021:

/metrics/domains/services/abc123?start=2021-09-01&end=2021-09-08&downsample=day&metric=requests

Returns the count of 200 and 204 status codes emitted by domains for service abc123, summed by minute, for the first 8 hours of September 2021:

/metrics/domains/services/abc123?start=2021-09-01T00:00:00&end=2021-09-02T08:00:00&downsample=minute&metric=status_200,status_201

Returns the count of 5xx status codes emitted by domains for service abc123 on the first day of September 2021, with one timeseries per unique domain:

/metrics/domains/services/abc123?start=2021-09-01&end=2021-09-02&metric=status_5xx&group_by=domain

Data latency

As a general rule, data with minute, hour, and day resolution will each be delayed by approximately 10 to 15 minutes due to the time it takes to collect and process historical data. Therefore, time periods that are yet to complete or which have only recently ended may return data that doesn't yet fully count all traffic for that period.

Error format

Error responses have the Content-Type header set to application/problem+json. The error response body will include:

• type (string): A URL that includes additional, human-readable information about this error type.
• title (string): A short name for the error type.
• status (integer): The HTTP status code.
• detail (string): A human-readable description of the specific error.
• errors (array): A list of individual errors, where multiple errors have occurred.

An example error can be seen below:

{
  "type": "https://fastly.help/metrics/validation-error",
  "title": "Request parameters were invalid.",
  "status": 400,
  "errors": [
    {
      "property": "metric",
      "reason": "Unrecognized metric names: 'my_metric', 'my_metric2'"
    }
  ],
  "detail": "Parameters with invalid values: 'metric'"
}

Domain Inspector provides a JSON-formatted historical API for visibility into responses delivered from your domains by Fastly. The API offers a standardized set of data to report in minutely, hourly, or daily granularity on every domain response, byte, and status code received by a specified service.

Data model

The unique combination of dimensions associated with this timeseries.

Entry data model

Values data model

The results of the query, optionally filtered and grouped over the requested timespan.

Endpoints